AI has solved one recruitment problem rather efficiently: finding the time to write another outreach message.

Give it a candidate profile and a job brief and, seconds later, you have something polished, professional and grammatically impeccable.

And quite possibly, completely forgettable.

The data backs up the need for a little caution.

Expandi, a LinkedIn outreach and automation platform used by sales teams, lead-generation agencies and recruiters, published its latest LinkedIn outreach benchmark analysis in May 2026, based on activity from May 2025 to April 2026. The study analysed more than 13.2 million LinkedIn connection requests and 6.7 million outbound messages across 13,302 active accounts. [source]

When Expandi compared AI-generated and human-written outreach within the same accounts, human-written messages achieved around 12% higher acceptance rates, while reply rates were broadly similar. [source]

That’s not a case for banning AI from the recruitment tech stack.

It’s a useful reminder about where AI adds value and where recruiter judgement still matters.

AI is excellent at the heavy lifting. It can research the candidate, summarise their career, identify relevant experience, spot connections to the brief and suggest different angles, draft the message.

Then the recruiter needs to step in and evaluate:

Why this person?

Why this role?

Why now?

Those three answers are what turn information into relevance.

AI can tell you that someone has spent five years building engineering teams.

The recruiter needs to know that the client is specifically looking for someone who has scaled an engineering function through international growth and that this particular candidate has done exactly that.

Now you have a reason to start a conversation.

That leads to a useful rule for recruitment teams:

Before AI writes the message, make the recruiter write the reason.

Ask the recruiter to complete:

“I’m contacting this person because ________, and I think they’ll care because ________.”

If the answer is:

“They have relevant experience and it’s a great opportunity.”

Back to the drawing board.

If it’s:

“She has just led a regional finance transformation, and our client needs someone who has done exactly that while integrating two acquired businesses.”

Now AI has something useful to work with.

There’s another interesting number in Expandi’s research: Staffing & Recruiting achieved an 18.9% message reply rate, compared with a 10.4% overall average. [source]

Recruitment outreach clearly can work. The opportunity here is to make the messages you send worth answering.

Think of the division of labour this way:

AI: research, organise, summarise, suggest, draft.

Recruiter: qualify, interpret, personalise, edit, decide.

The recruiter doesn’t need to beat AI at writing but to make sure AI is doing the work that saves time while the recruiter keeps the work that creates relevance.

Use AI to make your recruiters faster.

Just don’t let it make them sound interchangeable.

And if your entire team is sending beautifully written messages that could have gone to anyone…

you may have automated the wrong part.